Presiones e impactos derivados de las actividades del

This section describes corrections of inaccuracies related to scaling of inflow from season to year. In order to understand how such imprecisions might occur, it is useful to explain how the model handles scaling of seasons. From the description of the modeling framework in Section 3.1.2, it is known that for each year ten seasons are modeled, consisting of four regular and six peak- load seasons. These seasons are then scaled to comprise an entire year. This is done by converting the ten seasons into four main seasons, and scaling regular and peak-load seasons according to a predetermined weighting between the two types. Consequently, a parameter Ps that is defined for each season s is scaled to one year the following way:

Pyear = 8760 24⋅ 4⋅ 8750 8760

(

)

(

)

For the inflow parameters, where seasonal values are chosen based on the corresponding week the seasons reside in, the actual inflow value for this particular week will have a large impact on the total annual inflow. For example, if the inflow for the week that season 1 resides in is (incidentally) significantly larger than in the surrounding weeks, the inflow scaled to a year will be larger than the actual annual inflow. Also keeping in mind the relatively short season lengths it is understandable that scaling inaccuracies will occur. These imprecisions have to be taken into account and corrected for. Analyses for both the 3-scenario version and the 10-scenario version will therefore be performed.

5.5.1 The 3-Scenario Version 5.5.1.1 Regulated Inflow

Table 24 in Appendix D.1 gives country-wise aggregated regulated inflow for all three scenarios (aggregation is done in order to save space). The difference

is 81 TWh larger than the actual inflow for Norway. Evaluating scenario-wise differences, it is found that most of the deviation originates from scenario 2. For each of the scenarios, the scaled annual inflow used in the model is 115 TWh, 157 TWh and 103 TWh, respectively. Thus, inflow for scenario 2 is extremely high and should be corrected for. This is done in the next section. The variations for the other nodes and scenarios are generally not very large and does not have to be corrected for.

Correction of Norwegian Regulated Inflow

It is determined that the inflow is to be capped at 140 TWh (still a high value, but variations in scenarios are aspired). To do this, 17 TWh shall be removed from the total annual value. This removal is done in all ten seasons, based on their contribution towards the total annual inflow. The original inflow values, seasonal contributions and adjusted inflow are given in Table 25 in Appendix D.1. Utilizing the corrections outlined in the table and scaling the adjusted seasonal inflow values with Equation (5.8) indicates that the updated annual inflow is now correctly 140 TWh.

5.5.1.2 Run-of-the-River Inflow

The same analysis is carried out for run-of-the-river inflow. For all values, see Table 26 in Appendix D.1. Inflow is also here aggregated for all three scenarios. Most of the countries have small differences between the actual and scaled inflow. Norway has a difference of 25 TWh, or 18%, for the three scenarios combined. Further analysis demonstrates that most of the deviation again originates from scenario 2, with a variation of 16.2 TWh. This is a large difference and will be accounted for in the following.

Correction of Norwegian Run-of-the-River Inflow

Correction of the scaled inflow value in scenario 2 is done the same way as for regulated inflow. 16 TWh has to be removed from the scaled inflow to make up for the difference. The correction is done in all seasons, weighted towards each season’s inflow share. Table 27 in Appendix D.1 gives the original inflow, seasonal share of total inflow, the amount to be removed from each season and finally the adjusted inflow. Taking this removal into account and scaling the adjusted seasonal inflow values now gives an annual total inflow of 52.3 TWh for scenario 2. This is just above the scenario’s actual annual value of 49.5 TWh.

5.5.2 The 10-Scenario Version

The scaling of inflow from season to year is investigated the same way as for the 3-scenario version.

5.5.2.1 Regulated Inflow

The differences between actual and scaled regulated inflow are given as the sum for all ten scenarios in Table 28 in Appendix D.2. Poland has a very large difference. This is caused by the relatively low inflow values for this country, making the percentage-wise difference sensitive to changes. For most nodes, however, the summed difference for all ten scenarios can be considered negligible. Examining scenario-specific differences illustrates that the scenario- wise differences also are negligible for most nodes. Exceptions are Norway and Sweden. Here, the differences for each scenario are large enough that manual corrections have to be carried out.

Correction of Norwegian and Swedish Regulated Inflow

Table 29 in Appendix D.2 gives the actual inflow, scaled inflow and difference for each of the ten scenarios for Norway and Sweden. Regulated inflow for Norway and Sweden are corrected for by removing the percentage in the right-most column of the table. As with the 3-scenario version, the correction for each season is done depending on their share towards total annual inflow. The distribution of seasonal corrections is not presented in the table due to large amounts of data, but the procedure is the same as the one given for the 3-scenario version.

5.5.2.2 Run-of-the-River Inflow

Table 30 in Appendix D.2 gives a comparison of the actual and scaled inflow used in the model for the 10-scenario version. It can be seen that the country- wise differences for all scenarios is insignificant. Examining scenario-specific data, the largest difference for a single country (and all seasons) is found to be 35%, while the largest difference for each scenario summed over all countries and seasons is found to be 28%. However, the average difference is close to negligible. It is therefore decided to not perform any corrections of run-of-the- river inflow in this case.